Open RAN Cell-Site Capacity Forecasting for Proactive Congestion Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Wireless networks face challenges in efficiently managing resources and detecting congestion, particularly in rapidly-growing service areas, leading to poor response times, dropped calls, and suboptimal user experiences due to insufficient congestion detection mechanisms in cloud-based data and telephone networks.
Innovation Solution
A system that collects performance data from cell sites, determines busy-hour indicators, forecasts subscriber growth, and applies a gain function to detect capacity breaches, recommending capacity expansion to prevent future congestion.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If existing congestion detection mechanisms are used, then network resources can be monitored, but timely insights into evolving network conditions cannot be provided
Solution Approach 1:
The system performs preliminary actions by forecasting future network capacity requirements before actual congestion occurs. It uses historical performance data, subscriber growth models, and traffic patterns to predict future capacity needs, allowing network operators to proactively expand capacity before congestion impacts user experience.
Solution Approach 2:
The system dynamically adapts its detection and forecasting mechanisms to evolving network conditions. It continuously collects performance data, updates subscriber growth models, and adjusts capacity forecasts in real-time, enabling timely responses to changing network demands in rapidly-growing service areas.
2Productivity
If network capacity is increased to accommodate growing subscribers, then service coverage can be expanded, but network resources become overburdened
Solution Approach 1:
The system performs preliminary capacity planning by forecasting future capacity requirements before actual congestion occurs. It uses historical performance data, subscriber growth models, and traffic patterns to predict future capacity needs, allowing network operators to proactively expand capacity before congestion impacts user experience.
Solution Approach 2:
The system continuously collects network performance data and uses it to refine capacity forecasts. By monitoring actual traffic patterns, subscriber growth, and network utilization, it provides feedback loops that improve the accuracy of capacity predictions and enable data-driven capacity expansion decisions.
3Adaptability or versatility
If traditional RAN network techniques are used, then existing infrastructure can be leveraged, but cloud-based network specific challenges cannot be addressed
Solution Approach 1:
The system provides universal capacity forecasting capabilities that work across both traditional RAN networks and cloud-based networks. It collects performance data from diverse network types, applies unified forecasting models, and generates capacity recommendations applicable to different network architectures, making it versatile for various network implementations.
Solution Approach 2:
The system adapts its analysis parameters and forecasting models to match the specific characteristics of cloud-based networks. It adjusts capacity thresholds, performance metrics, and growth models to account for cloud infrastructure differences, enabling accurate capacity forecasting for cloud-native network architectures.
Data Source
AI summary
An example process may collect performance data from a cell site of an open radio access network (RAN) in an area of interest (AOI). The cell site may include sectors and the sectors comprising cells. Busy-hour indicators may be determined based on the performance data from the cell site. The busy-hour indicators may be determined by applying a percentile method to outliers in the performance data. A subscriber growth model in the AOI can be forecast for a forecast period. The busy-hour indicators can be extrapolated using the subscriber growth model to generate forecast indicators for the forecast period. A gain function can be applied to the forecast indicators to generate revised forecast indicators. A capacity breach at a cell or a sector of the cell site in the AOI can be detected in response to an indicator from the revised forecast indicators exceeding a capacity threshold.